evidoria

← Back to browse

Good practice Imported

Gemini AI In-Class Support Outperforms Standard Instruction for Omani EFL Grammar Learners

Oman · Sohar · See the Oman profile

Evidence: Quasi-experimental Top 59% 47/100 · Ask Evidence Copilot about this practice

A controlled study of 75 Omani pre-intermediate EFL learners found Gemini AI used inside the classroom produced the largest grammar gains and highest engagement compared with AI used outside class or standard instruction alone, published in Computers & Education Open (2025).

n=75 (3 groups of 25)
Study sample size
Gemini AI In-Class Support Outperforms Standard Instruction for Omani EFL Grammar Learners

Details

Maturity
Pilot
Promoter
Sohar University / University of Technology and Applied Sciences, Shinas
Period
2025
Keywords
EFL, language learning, generative AI, higher education

Context

Oman's higher-education English-language programmes have been exploring generative AI tools to support grammar instruction, but evidence on how classroom integration compares with unsupervised use had been limited.

Objectives

The study set out to compare whether Gemini AI used inside the classroom, Gemini AI used outside class hours, or traditional in-class instruction alone produced the largest grammar gains and engagement for pre-intermediate Omani EFL learners.

Activities

Researchers divided 75 Omani pre-intermediate EFL learners into three equal groups of 25 — a control group with traditional in-class instruction only, an experimental group using Gemini AI within the classroom alongside traditional instruction, and a second experimental group using Gemini AI outside class hours — and administered researcher-designed, piloted grammar pretests and posttests to all three.

Results

All three groups improved from pretest to posttest, but the in-class Gemini AI group significantly outperformed both the out-of-class AI group and the control group on grammar gains, and also reported the highest engagement across the measures used in the study.

Conclusions

The published account reports the direction and statistical significance of the results but does not disclose exact effect sizes, and the single-institution, single-skill design means the classroom-integration advantage may not generalise beyond grammar instruction for pre-intermediate EFL learners.

Implementation
Indicative cost
Low (< €50k)
Time to results
Short (< 1 year)
Staffing & skills
Instructors integrating Gemini AI into classroom sessions, Researchers administering pretests, posttests and engagement measures

Conditions for success

  • In-class facilitation and supervision when using the AI tool, rather than unsupervised use
  • Piloted, researcher-designed grammar assessments to measure gains consistently across groups

Common failure modes

  • Exact effect sizes are not published, limiting comparability with other interventions
  • The single-institution, single-skill (grammar) design limits generalisability

Commonly funded by

National / regional programmes Own resources / municipal budget

Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.

Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.

Data sources

Where this practice's information was retrieved from, and when.

Attachments

Similar practices you may find useful